The present invention relates to image forming apparatuses, methods for forming an image, and computer-readable recording media. In particular, the present invention relates to techniques for reducing the time required for the entire calibration process.
In an image forming apparatus, such as a color printer or a color multifunction peripheral, the electrical and mechanical conditions that are required for image forming and output operations (color printing) are modified in accordance with changes in the environment where the image forming apparatus is used, the level of wear and tear on the components, the number of printing operations, etc. For example, when color printing based on the same image data is performed on different days, the color and density of an image on the first printed sheet may be different from the image on the second printed sheet, due to changes in the electrical and mechanical conditions described above.
As a solution to this issue, an image forming apparatus with color printing capability performs a calibration that involves correcting color or density for to resolve the problem of color change or density reduction in printed images (output images). Execution of such a calibration makes it possible for the output images on the first and second printed sheets to have consistent image quality.
There are several types of calibrations, including bias calibration, I/O calibration, and registration calibration. Bias calibration corrects a bias (developing bias) applied to a developing device (developing roller) in accordance with the density of a test image (which may hereinafter be referred to as a correction image or a patch). I/O calibration corrects a color density gradient (which may hereinafter be referred to as a γ table) used to correct the color density of an actually formed image (output density) relative to the density of a predetermined color in image data (input density). Registration calibration measures the position of a patch formed in a predetermined shape and corrects misregistration of the patch. For example, predetermined types of calibrations are performed depending on the specifications, settings, or usage of the image forming apparatus.
Conventionally, these three types of calibrations; i.e., bias calibration, I/O calibration, and registration calibration have been performed sequentially at a predetermined time, such as when the image forming apparatus is turned on or when a predetermined number of printed sheets have been outputted.
A series of calibrations that are conventionally performed will now be described.
FIG. 8 illustrates an example of patch patterns that are used in a series of calibrations in the prior art. The different sections illustrated in FIG. 8 correspond to respective three turns of an intermediate transfer belt B1.
In a series of calibrations that are conventionally performed, first, a bias is corrected by executing a bias calibration. Then, an I/O calibration and registration calibration are performed using the corrected bias.
That is, in the series of calibrations, as illustrated in FIG. 8, in section 801 for the first turn of the intermediate transfer belt B1, a background density at a position for forming a patch pattern (a predetermined number of patches) 800a for bias calibration and a background density at a position for forming a patch pattern 800b for I/O calibration are calculated (measured) using two density sensors 802 and 803, respectively.
Next, in section 804 for the second turn of the intermediate transfer belt B1, the patch pattern 800a for bias calibration is formed using a predetermined bias, at a position corresponding to the position at which the background density was measured. Then, the predetermined bias is corrected based on the density (measured density) of the patch pattern 800a and the density (target density) for forming the patch pattern 800a. 
Next, in section 805 for the third turn of the intermediate transfer belt B1, the patch pattern 800b for I/O calibration and a patch pattern 800c for registration calibration are sequentially formed at predetermined positions using the corrected bias. Then, a γ table and misalignment are corrected using the patch pattern 800b and the patch pattern 800c, respectively.
However, in the series of calibrations in the prior art, as illustrated in FIG. 8, the bias calibration needs to be completed before execution of the I/O calibration and registration calibration that require a corrected bias. To complete the bias calibration, it is necessary that a patch at the trailing end of the patch pattern 800a (in the running direction of the intermediate transfer belt B1) reaches a predetermined detectable range of the density sensor 802 so that its density can be detected. Therefore, during the period from formation of the patch pattern 800a for the bias calibration on the intermediate transfer belt B1 until detection of the density of the patch at the trailing end of the patch pattern 800a, it is not possible to form, on the intermediate transfer belt B1, the patch pattern 800b for I/O calibration and the patch pattern 800c for registration calibration.
As a result, on the intermediate transfer belt B1, an empty space 806, where no patch pattern is formed, is created immediately behind the patch pattern 800a, as illustrated in FIG. 8. This means that the time required for the series of calibrations (i.e., the entire calibration process) increases by the amount of time that corresponds to the empty space 806.
The bias that influences the color or density of the image on a printed sheet changes depending on predetermined factors, such as temperature and humidity within the image forming apparatus. However, when bias calibration is frequently performed, even if a bias is corrected in the bias calibration, there may be no significant difference between the uncorrected bias and the corrected bias. For example, when the image forming apparatus is repeatedly turned on and off in a short period of time for maintenance operation, a bias in the previous bias calibration and the most recent bias calibration are substantially the same. In such a situation, it will not be necessary to wait for the result of bias correction in bias calibration before forming the patch pattern 800b for I/O calibration and the patch pattern 800c for registration calibration.
In recent years, a predetermined number of parameters that influence changes in bias have been discovered. This means that by determining these parameters, it is becoming possible to estimate (determine) a variation in bias under conditions of the determined parameters, based on the previously calculated relationships between bias and the predetermined number of parameters. In other words, that a difference between a bias estimated from past data (estimated value) and a bias obtained by bias calibration actually executed (measured value) is decreasing. If a corrected bias can be accurately estimated from past data and a predetermined number of parameters, there is no problem in forming the patch pattern 800b for I/O calibration and the patch pattern 800c for registration calibration using the estimated bias, without waiting for the result of bias correction in bias calibration. The time required for the entire calibration process can thus be reduced, which was not achievable in the prior art.